It was only a game. But this 2023 exercise conducted in Washington illustrated just how easily the current global crises could escalate into nuclear war.

So where might a nuclear war start exactly? What would it look like? And how many of us would survive?

There are many potential flashpoints. 

This week, we were reminded of the perils of escalation when India and Pakistan launched tit-for-tat air, missile and drone attacks following the deadly terrorist attack in Indian-controlled Kashmir. 

The conflict so far is a conventional one. But both countries also have sizeable nuclear arsenals. 

In 2022, the region momentarily pirouetted on threshold of oblivion when India accidentally fired a cruise missile at Pakistan (it was unarmed; no one was hurt).

Indeed, before the full-scale Russian invasion of Ukraine, most arms control experts considered just such a flare-up between India and Pakistan as the most likely starting point for nuclear apocalypse.

Then there is the Middle East. 

This weekend, American and Iranian envoys will sit down in Oman for their fourth round of talks on Tehran’s nuclear programme. Failure could result in a US-Israeli strike like the one described above.

In the Pacific, meanwhile, China is projected to nearly double its nuclear arsenal to 1,000 warheads by 2030. In North Korea, Kim Jong-un says he wants to expand “exponentially” his stockpiles. 

The journalist Annie Jacobsen’s recent book Nuclear War: A Scenario traces in terrifying detail how North Korea could start a global war by firing a single rocket at the United States.

And in Europe, Vladimir Putin has repeatedly rattled the nuclear sabre. Last November he fired a nuclear-capable ballistic missile at the Ukrainian city of Dnipro: a blunt declaration that Cold War brinksmanship is back. 

This is the climate in which the United Nations General Assembly has voted to establish an international panel of scientists to assess, communicate and advance our current knowledge of the effects of nuclear war.

The investigation, set up in autumn last year and due to report in 2027, will look at “the physical effects and societal consequences of a nuclear war on a local, regional and planetary scale, including, inter alia, the climatic, environmental and radiological effects, and their impacts on public health, global socioeconomic systems, agriculture and ecosystems, in the days, weeks and decades following a nuclear war”.

In the Cold War, theories of nuclear conflict focused overwhelmingly on a world-ending holocaust involving the arsenals of the United States and the Soviet Union.

Today, there are many more players, but the logic of destruction is unchanged.

“You start with a local conflict somewhere, it eventually moves up the escalation ladder, and one side or the other eventually takes the steps to use a battlefield nuclear weapon, and the other side decides to respond,” says Alexander Bollfrass, the head of strategy, technology and arms control at the International Institute for Strategic Studies.

That is so well-rehearsed a scenario that complex mechanisms have been developed to prevent it.

Russia warned the US in advance about the intermediate-range strike that hit Dnipro in November via what arms control wonks call “nuclear deconfliction channels”.

Much of this mechanism, like the hotline set up between the White House and the Kremlin following the Cuban Missile Crisis, was built by politicians and deterrence theorists following incredibly scary close calls.

But not every country has had the time to develop that kind of nuclear strategic culture, which is one reason proliferation is so frightening.  

The case for going to war to stop Iran or other countries from getting a nuclear weapon, one Western official speaking on condition of anonymity tells The Telegraph, is much stronger when you realise the world only survived the 1950s and 1960s through “dumb luck”.

Similar concerns have been raised over the years about other countries joining the nuclear club, Bollfras points out.

So far, however, Israel, India, Pakistan, and even North Korea have all managed to restrain themselves from using them in anger, and manage relations with their adversaries.

“But I completely agree with the  broader point,” Bollfras adds. “In my academic life, I spent a lot of time looking at nuclear dynamics during the Cold War, and I came out of that experience just not understanding how it is that we’re all still alive. We’ve been incredibly fortunate so far.

“There are quite a few studies in the nuclear weapons field about the nature of the nuclear taboo, because it is really quite remarkable that nuclear weapons have only been used in anger by one country, a very long time ago. And there’s a lot of competing explanations out there, a lot of competing beliefs, about why that is.”

If – or when – that taboo fails, things will move incredibly fast.

American or Russian early warning systems will know in an instant that an intercontinental ballistic missile (ICBM) has been launched. But they may not be able to tell where it is flying for several minutes.

By the time someone wakes up the president or prime minister, he or she will have less than 10 minutes to make a decision about firing a response.

And at that point, “no one has an idea of what will happen,” says Victor Gilinsky, a physicist who acts as programme advisor to the Nonproliferation Policy Education Center and was a commissioner of the US Nuclear Regulatory Commission during the Ford, Carter and Reagan administrations.

“None of the hundreds of books on deterrence, et cetera, are of any use. It will come down to two politicians who haven’t read them, haven’t taken the Harvard course on deterrence, probably haven’t slept in a crisis, and may be on drugs to stay alert. It will be like a contest between two bullies in a schoolyard. They might walk away from a fight, and they might kill each other.

“Their advisers will also want to look tough. I don’t think anyone wants a nuclear war, but the shadow-boxing could easily get out of hand. One side may feel risking nuclear war is preferable to being humiliated, the most awful thing for a typical leader.

“As to what happens if it comes to nuclear weapons, that depends totally on how they are used.”

The Arms Control Association, a US-based nonpartisan organisation, counts more than 12,100 nuclear weapons in the world today, 90 per cent of them held by the US and Russia. China, France, Britain, India, Pakistan, Israel and North Korea hold the remainder.

They are generally delivered by very fast ICBMs that fly into space and can reach targets less than half an hour after launch. Some carry multiple warheads that then independently plummet to different targets. They are almost impossible to intercept.

While we tend to think of mushroom clouds over Hiroshima and Nagasaki and memories of GCSE physics – the banging together of lumps of uranium or plutonium until electrons start smashing apart atoms – most of the weapons in the nose cones of today’s missiles are different beasts entirely.

These “thermonuclear” weapons use a Hiroshima-style device only as the trigger for another device, nestled next to it in the same bomb casing, that contains isotopes of hydrogen that fuse into helium atoms.

That is the same thing that happens in the centre of the sun, and it generates extraordinary amounts of energy.

The Hiroshima bomb killed 66,000 people with 15 kilotons of power - the equivalent of 15,000 tons of conventional TNT high explosive going off at once.

But the most powerful thermonuclear weapons ever exploded, the US Castle Bravo test in 1954 and the Soviet Tsar Bomba in 1961, released 15 and 50 megatons respectively.  

One megaton equals 1,000 kilotons. There is, Soviet scientists concluded after testing the Tsar Bomba, practically no upper limit to the destructive power of such devices. 

Thanks to more than 2,000 nuclear weapons tests since 1945, we know an awful lot about what happens when you set off one of these things.

Vast amounts of research have gone into how the altitude of the explosion, the power (“yield”, in nuclear jargon) of the blast, the weather conditions, and other variables equate to damage, radiation and human casualties.

Much (presumably not all) of that data has been made public, and the nuclear scientist Alex Wellerstein has done the world a macabre favour by collating it into an online tool called Nukemap, which allows you to simulate the impact of various kinds of in-service warheads anywhere in the world.

The one bit of good news: detonation at that altitude means the fireball would probably not touch the ground, so radioactive fallout is minimal.

Reduce the altitude at which the warhead explodes by a few hundred yards, however, and prevailing winds carry a plume of radioactive filth over East Anglia.

But that is just the immediate effect.

“The military knows a lot about blast and about radioactivity, but they don’t reliably model fires, and especially the indirect effects, the effects of the smoke and effects on climate,” says Prof Alan Robock of Rutgers University in New Jersey.

“We’ve calculated that 10 times as many people would die around the world from these indirect effects than would die from the horrific direct effects that we already know about. And they’re trying to ignore this because they want to continue business as usual.”

Prof Robock is not a physicist. He is a meteorologist, and 40 years ago he was among a small number of American and Soviet researchers who began to consider the impact of mass nuclear explosions on the weather.

Their conclusions were apocalyptic.

Nuclear attacks on cities and industrial areas would start fires and pump smoke and dust high into the stratosphere, they realised. Because that is high above the clouds, there would be no rain to wash it down again, meaning it would remain there for years.

This layer of floating dust would absorb sunlight, making conditions colder and darker on earth’s surface. Less sunlight means less evaporation from the oceans, which means less rain, so it would also become much drier.

A nuclear war between the US and Russia “could produce enough smoke to make temperatures go below freezing in the middle of continents, even in the summertime”, Prof Robock says.

“Clearly that would completely shut down agriculture. The world has about a 60-day food supply. So once people ate up all the food that’s been stored, they’d have to start growing it again. So we calculated, at the end of the second year, how much food could each country produce?

“We looked at every country: what do they eat? How much do they export? How much do they import? And we calculated a war between India and Pakistan could kill one to two billion people from starvation. A war between the US and Russia and Nato could kill six billion people, most of the people on the earth. That’s what nuclear winter is.”

Nuclear winter quickly entered the popular imagination.

Ronald Reagan and Mikhail Gorbachev cited it when they declared that “a nuclear war cannot be won and must never be fought” and agreed to move towards disarmament.

The idea has not been without its critics. Some argue that the assumptions are arbitrary and ignore the dozens of variables – from weather, to target, bomb size, to altitude of explosion – that could produce different outcomes.

But Prof Robock says he and his colleagues have tested the hypothesis using the most advanced computer modelling developed by climate change scientists in recent years, taking data from forest fires, volcanic eruptions and similar events as analogues.

It still stands up, he says. And for the governments of the nuclear-armed powers, he suspects, that is not welcome news.

“If there is a nuclear winter, their whole philosophy of deterrence falls apart, and it implies that we have to get rid of all the nuclear weapons,” he says. “And there are huge forces in nuclear countries whose careers and earnings depend on maintaining the status quo”

In Prof Robock’s models, some countries, mostly in the southern hemisphere, are less affected than others. He mentions Australia and Argentina among the (relatively) lucky ones.

But what about the rest of us? There are plans in place for what happens should a warhead hit Britain.

Authorities maintain a Mass Evacuation Framework for London, which is meant to be implemented in various emergencies and envisages pre-planned routes for getting people out of the capital.

In the former Soviet Union, many factories and other institutions have bomb shelters in their basements, although they would be of questionable use against a direct hit. The metro systems in places such as Moscow and Kyiv are designed to double as mass refuges.

But in reality, no one seems to have much of a real plan – at least not for the general public.

Some years ago, Bollfrass was leafing through the archives of the Stasi, Communist East Germany’s secret police service, when he came across a study of nuclear war.

“The interesting thing,” he says, “was that they acknowledged to their own intelligence officers that they were in no way prepared for it, and that in the event of nuclear war, there’d be a few bunkers available to the leadership, but everyone else will be on their own. Which at least was refreshingly honest on their part.

“I don’t think that Western governments would be quite as candid, although I think back then and even today, the situation isn’t really all that much better.”

However, no number of bunkers can counter the profound knock-on effects on climate, economy, trade and social cohesion that the proposed United Nations study is meant to look at.

So, for now, the basic plan is simply to make sure nuclear war does not happen.

Some fear that the taboo that has held since 1945 is slowly weakening.

That is partly just because, with more wars in the world, more people are discussing the possibility of nuclear use.

Partly, too, because theories such as nuclear winter cannot be tested for real, meaning there has always been room to argue that it would not be as bad as all that.

And partly because there is a temptation to think modern weapons’ designs could minimise the chances of global catastrophe.

More accurate missiles mean less chance of missing, and less need for a massive explosive power to ensure hitting a target. Wiping out a city is not necessary.

By carefully choosing whether to detonate in the air, on ground impact, or underground (such as with a bunker-busting warhead), the amount of debris and fallout could be minimised.

And the established nuclear powers have replaced their multi-megaton weapons of the Cold War with “dial-a-yield” weapons that commanders can set to go off with varying force.

The Holbrook warheads on Britain’s Trident missiles, for example, are thought (the details are secret) to vary from 100 kilotons all the way down to 0.3 (about three times the 2020 Beirut port blast).

Put these sensible precautions together, and you can easily reason yourself to a place where letting off a “small” nuclear bomb, aimed at a purely military or even entirely uninhabited target, might not be as unthinkable as you initially assumed.

People who study nuclear weapons insist that would be madness. But the hope of being able to use such warheads in a “restrained” way may have been on display when the players representing Israel in Sikorski’s 2023 wargame opted initially for a blast over the Iranian desert.

When that failed to intimidate their opponent, they deliberately chose military targets and avoided bombing Tehran or other cities. The Iranian players similarly chose not to nuke Tel Aviv.

That doesn’t mean they would not have targeted population centres if the game had continued, or that real prime ministers and generals would make the same decisions in real life.

But all of those factors, says Sikorski, makes it almost impossible to predict what the outcome of an actual nuclear war would be.

One of the noticeable things about the Iran-Israel simulation, he says, was that the players – 35 Republican and Democratic Capitol Hill staff, American government officials and analysts, academics, military personnel, and national security and Middle East think-tankers – found it difficult to develop appropriate policy responses to Israeli or Iranian nuclear use. They simply hadn’t ever planned or been trained for such a scenario.

In other words, the players had spent so much time hoping deterrence would work that they hadn’t thought about what to do if it failed.

More exercises forcing officials to grapple with the profound dangers ought to help them plan ways through such crises if they encounter them. Above all, it would act as a vaccine against complacency.  

There is a fine line between preparing for the worst and unleashing paranoia and panic. But, Bollfrass says, it is a tightrope that must be walked.

“We’ve all had experiences where we spend our professional lives thinking about these issues, and when we think there might be a moment where there’s cause for concern, we become frustrated that we can’t motivate people to engage with it,” he says.

“At the same time, certainly in Europe after the full-scale invasion [of Ukraine], there have been times where it seemed like public fears and media reporting seemed to create an atmosphere of panic, where we [the small community of nuclear experts] kind of feel it’s our role to step in and say, ‘No, no, everyone, calm down.’”

Sikorski, for his part, would like to see more officials reckoning with the dilemmas exposed by his wargame.

“I think it is dangerous to be flippant about the dangers of using nuclear weapons, or the dangers of possessing them.”